Validity of Otoliths and Pectoral Spines for Estimating Ages of Channel Catfish
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
North American Journal of Fisheries Management 22:934–942, 2002 q Copyright by the American Fisheries Society 2002 Validity of Otoliths and Pectoral Spines for Estimating Ages of Channel Catfish DAVID L. BUCKMEIER* Texas Parks and Wildlife, Heart of the Hills Research Station, HC 7, Box 62, Ingram, Texas 78025, USA ELISE R. IRWIN U.S. Geological Survey, Alabama Cooperative Fish and Wildlife Research Unit, 1 108M White Smith Hall, Auburn, Alabama 36849, USA ROBERT K. BETSILL AND JOHN A. PRENTICE Texas Parks and Wildlife, Heart of the Hills Research Station, HC 7, Box 62, Ingram, Texas 78025, USA Abstract.—Basal recess and articulating process sec- estimate the age of channel catfish Ictalurus punc- tions of pectoral spines are often used to estimate the tatus and flathead catfish Pylodictis olivaris; how- age of channel catfish Ictalurus punctatus. However, ever, interpretation of annuli can be difficult identification of annuli in pectoral spine sections can be difficult. We developed and validated a method for es- (Sneed 1951; Marzolf 1955; Mayhew 1969; Pren- timating the age of channel catfish by using sagittal oto- tice and Whiteside 1975; Turner 1982; Crumpton liths. We also validated a new method using pectoral et al. 1987). As fish age, expansion of the central spines in which a single cut is made through the dorsal lumen erodes early annuli, thereby causing the true and anterior processes (hereafter termed cut spines) and age of older fish to be underestimated (Muncy annuli are enhanced with side illumination. Age esti- 1959; Mayhew 1969). Section location is also im- mates from otoliths and cut spines were compared with age estimates from traditional articulating process sec- portant; for example, sections through the basal tions of pectoral spines for channel catfish of known recess often have fewer annuli than would match ages (1–4). Age estimates by the three methods were the true age of flathead catfish (Turner 1982; Nash correct for more than 90% of fish after two experienced and Irwin 1999). Turner (1982) noted that some readers independently estimated ages and resolved dis- annuli in spines are composed of multiple growth agreements by mutual examination. Otoliths were more rings, and Sneed (1951) noted that the irregular accurate and less variable in estimating age. Otolith age estimates were always within 1 year and, after mutual arrangement of growth rings presented some dif- examination of the structures in question, 97% of the ficulties. False marks are usually distinguishable assigned ages agreed with known age. The accuracy of in younger fish because of their proximity to true cut spines and articulating process sections after mutual annuli (Turner 1982); in older fish, however, false examination was similar; however, the cut-spine method marks may become more problematic and lead to was simpler than preparing articulating process sections. overestimation of fish age. Finally, in slow grow- Otolith annuli were more distinguishable than pectoral ing and old individuals, spine annuli near the edge spine annuli and were validated for age-1–4 channel catfish. Therefore, we recommend using otoliths to es- tend to merge and may be indistinguishable (Lai timate the age of channel catfish. et al. 1996; Kocovsky and Carline 2000), thereby increasing the chance of biased age estimates. Otoliths are valid structures for estimating ages Historically, basal recess or articulating process of many freshwater fish species and often are pre- sections of the pectoral spine have been used to ferred over spines and scales (Taubert and Tran- quilli 1982; Erickson 1983; Heidinger and Clod- * Corresponding author: daveb@ktc.com felter 1987; Hall 1991; Hales and Belk 1992; Hin- 1 Cooperators are U.S. Geological Survey; Alabama ing et al. 2000). However, the use of otoliths for Department of Conservation and Natural Resources, estimating ages of adult catfishes has been limited. Game and Fish Division; the Alabama Agricultural Ex- Crumpton et al. (1987) reported that otolith annuli periment Station; and Auburn University. were difficult to distinguish in channel catfish, Received June 19, 2001; accepted November 11, 2001 brown bullhead Ameiurus nebulosus, and white 934
MANAGEMENT BRIEFS 935 FIGURE 1.—Location of cut used to remove otoliths from channel catfish. The cut (dotted line) is made about 3–5 mm anterior to a line that would connect the locked pectoral spines (solid line). catfish A. catus. However, Nash and Irwin (1999) were collected and removed from the pond in found otoliths to provide an efficient and accurate March, April, May, June, and October 1998 to help method for estimating ages of adult flathead cat- determine time of annulus formation. Additional fish. collections were made in July 1999 and 2000. Fish Population age structure and growth are mea- collected represented known-age (1–4) channel sures commonly used to manage fisheries. How- catfish. ever, many studies do not consider the accuracy Channel catfish were collected from the pond of their age estimates (Beamish and McFarlane by electrofishing and angling. Collection date and 1983). Erroneous age estimates may affect man- total length (mm) were recorded for each fish. Sag- agement decisions. Therefore, the relative accu- ittal otoliths were removed by sectioning through racy of each method should be considered when the supraoccipital bone using a hacksaw. The cut choosing a method to estimate age. Our objective was made about 3–5 mm anterior to a line that was to develop and validate improved methods for would connect the locked pectoral spines (Figure estimating the age of channel catfish by using oto- 1). Otoliths were then removed from the posterior liths and pectoral spines. portion of the skull by forceps. If otoliths were difficult to locate, the top of the skull was removed Methods with wire cutters to provide better access. Otoliths Channel catfish (N 5 225) from the 1996 year- were stored and dried in vials marked with ref- class were acquired from the Alabama Agricultural erence numbers. Pectoral spines were disarticu- Experiment Station, Auburn University, in January lated and removed (Mayhew 1969). Spines were 1998. Sixteen fish were sacrificed to provide a ref- cleaned and stored in coin envelopes marked with erence collection of bony structures for estimating reference numbers. age. Adipose fins of the remaining fish were Because otoliths had not yet been successfully clipped and the fish were stocked into a 0.57-ha used to estimate the age channel catfish, we ex- pond located at Auburn University. Forage was perimented with several processing methods—in- available in the form of a previously established cluding those of Maceina (1988), which were used sunfish Lepomis spp. population. Channel catfish by Nash and Irwin (1999) to age flathead catfish,
936 BUCKMEIER ET AL. and those suggested by Bagenal and Tesch (1978), each method were assessed by comparing actual which were used by Heidinger and Clodfelter mean length at age and age structure with the es- (1987) to age smallmouth bass Micropterus do- timates by each reader and those following a con- lomieu, walleye Stizostedion vitreum, and striped cert reading. Actual mean length-at-age was de- bass Morone saxatilis. Otoliths were not used to termined by calculating the mean of the individual estimate ages of all fish collected because some lengths at capture for each age-class. Age structure otoliths were destroyed during method develop- was the number of fish of known age collected ment. from each age-class. Mean length-at-age and age We experimented with alternative processing structure estimates, by each reader and following methods for pectoral spines to reduce the influence a concert reading, incorporated age estimation er- of the central lumen and to improve the recognition ror. Comparisons of estimates and actual values of annuli. Ideal section location was determined would reveal whether age estimation error could by cutting multiple cross sections (75 mm thick) potentially alter management decisions. Correct of several pectoral spines with a Buehler low- age assignments for all fish in a sample would thus speed isomet saw. Sections for each spine were yield estimates of mean length-at-age and age mounted on microscope slides and viewed through structure that were identical to actual values. a dissecting microscope (5–253). Each section was examined and the number of annuli and the Results relative size of the central lumen were noted. Once Sample Preparation the ideal location was determined, we experi- mented with side illumination (similar to the meth- Annuli were not consistently visible in channel od for otoliths described by Heidinger and Clod- catfish otolith sections prepared by the methods of felter 1987) to improve annulus recognition. Maceina (1988), but they were visible in channel We prepared and examined articulating process catfish otoliths prepared and viewed by methods sections of pectoral spines (similar to the method similar to those of Heidinger and Clodfelter described by Turner 1982) to compare age esti- (1987). Some modifications were necessary be- mates from articulating process sections with age cause annuli were difficult to discern. Burning oto- estimates from otoliths and pectoral spines. Artic- liths by placing them directly on a hotplate (Corn- ulating process sections were viewed by using ing PC 351; Corning Industries, Corning, New transmitted light and a dissecting microscope (5– York) changed their color from opaque white to 253) for channel catfish ages 1–3. yellowish brown (after 30–60 s at medium heat) For each structure and preparation method, and enhanced annular marks. Mounting each spines or otoliths from fish of unknown age were burned otolith on a microscope slide by using an combined with structures of known age to reduce adhesive (Crystalbond; Buehler, Lake Bluff, Illi- bias. Two readers, each with numerous years of nois) allowed easier processing, given the very experience estimating fish age, independently es- small size of channel catfish otoliths (Figure 2). timated the ages of structures without reference to To ameliorate the brittleness of channel catfish oto- fish length. Disagreements were resolved by mu- liths, support was provided by forming adhesive tual examination of the questionable structures around the lower half of the otolith. Otoliths were (hereafter termed a concert reading). Spines and mounted with the rounded anterior edge upright otoliths from fish of unknown age were not in- and the pointed posterior edge in contact with the cluded in analyses. Percentage reader agreement microscope slide. The anterior–posterior axis was was calculated for each method to compare pre- perpendicular to the slide; otherwise, visibility of cision. The percentage of correctly estimated fish annuli decreased. Once the adhesive hardened ages was calculated for each method to compare (30–60 s), we hand-sanded the otolith with wetted accuracy. Age bias graphs (similar to those of 600-grit carborundum sandpaper until 1/3–1/2 of Campana et al. 1995) were generated for each the otolith was sanded away revealing the nucleus method of estimating age and were used to assess in the transverse plane (Figure 2). Coating the individual reader bias. In addition, age bias graphs sanded surface with mineral oil and using side were constructed after a concert reading to deter- illumination enhanced the appearance of annuli as mine whether the use of a concert reading cor- viewed with a dissecting microscope (10–603). rected any individual bias associated with each We found that side illumination with a 1-mm- method. diameter, single-strand fiber optic filament con- The implications of the error associated with nected to a light source (Fiber-Lite model 180;
MANAGEMENT BRIEFS 937 FIGURE 2.—(A) Illustration of the right otolith from a channel catfish depicting how it is mounted to a glass slide. The otolith is sanded from the anterior tip to the dashed line, resulting in the transverse view through the nucleus shown in the photograph. (B) The photograph of an otolith is from a 4-year-old channel catfish shows the location of the nucleus (N) and annuli (circled); the fourth annulus is located at the edge. Dolan-Jenner Industries, Inc., St. Lawrence, Mas- For pectoral spines, a single cut through the dor- sachusetts) provided the best results. Reflecting sal and anterior processes (Figure 3) provided the light at several angles, including through the oto- best age estimates with the least effort. Serial sec- lith and at the sanded surface, enabled observation tions of spines revealed that as early as age 1, of all annuli. annuli were missing in some basal recess sections.
938 BUCKMEIER ET AL. FIGURE 3.—Location of the cut (dotted line) used to process cut pectoral spines for estimating the age of channel catfish. The saw blade should be aligned from the tip of the ventral process (V) to the outer edge of the dorsal process (D) and section through the dorsal and anterior (A) processes. This location minimizes the influence of the central lumen. The central lumen was smallest in a section taken within 1 year of known age (Figure 4). Following through the dorsal and anterior processes. We concert readings, only one age estimate was in- found that this cut could easily be made by mount- correct; an age-4 fish was estimated to be age 5. ing a spine in a small hobby vise with rubber jaws Agreement between readers was 79%. Both read- and cutting the spine with a hand-held jeweler’s ers indicated the first annulus was the most difficult saw equipped with a 5/0 blade. Embedding the to detect. distal end of the spine in clay and coating the cut Cut-spine and articulating process section age surface with mineral oil enhanced visibility of an- estimates were more variable between readers than nuli when viewed through a dissecting microscope were otolith age estimates (Table 1). Agreement (5–253). As with otoliths, a 1-mm fiber optic fil- between readers was slightly better for cut spines ament provided effective side illumination. Here- (65%) than for articulating process sections (52%). after this procedure is termed the cut-spine meth- Reader error ranged from 1 to 6 years (Figure 4). od. The first reader was consistently less accurate (Ta- ble 1), even though both readers were experienced Validation at estimating fish age. After concert readings, how- Annulus formation started in late April and was ever, age estimates from cut spines and articulating complete by early June. Otoliths provided the most process sections were within 2 years of the known accurate age estimates of 1–4-year-old channel age, and more than 90% were correctly estimated. catfish for each reader and following a concert With a concert reading, ages estimated from cut reading (Table 1). The first and second readers spines and articulating process sections were ad- correctly estimated the age of 82% and 95% of the equate for 1–3-year-old channel catfish. However, otoliths, respectively. Otoliths in which age was age estimates of age-4 fish exceeded the known improperly estimated by the readers were always age by 1–2 years.
MANAGEMENT BRIEFS 939 TABLE 1.—Percentages of otoliths, cut pectoral spines, as based on pectoral spines. However, individual and articulating process (AP) sections for which ages were interpretation of annuli was more variable for pec- estimated correctly by each reader and following a concert toral spines. Prentice and Whiteside (1975) also reading, as well as the precision (percent agreement be- tween readers) for known-age channel catfish (ages 1–4). noted much variability between readers who were estimating ages of channel catfish spine sections. Accuracy Conversely, reader agreement for otolith age as- Concert signments was high, indicating annuli on otoliths Age N Reader 1 Reader 2 reading Precision were easier to identify. Annuli on otoliths have Otoliths also been determined to be more distinguishable 1 13 85 92 100 85 than spine annuli for flathead catfish (Nash and 2 9 67 89 100 56 Irwin 1999) and walleye (Erickson 1983; Kocov- 3 14 100 100 100 100 4 3 33 100 67 33 sky and Carline 2000). Although channel catfish Overall 39 82 95 97 79 otolith annuli were distinct, they were generally Cut spines lighter than in other fishes such as largemouth bass 1 28 68 100 100 68 Micropterus salmoides, crappie Pomoxis spp., and 2 18 89 94 89 83 flathead catfish. With practice, annuli on channel 3 16 38 94 94 38 catfish otoliths were easily identified. 4 3 0 0 0 67 Overall 65 63 94 91 65 The error from compounding sources probably AP sections makes spines an improper choice for estimating 1 28 68 96 100a 64 ages of old channel catfish, especially in slow- 2 18 72 83 94 56 growing populations. We found that annular marks 3 16 25 88 81 25 composed of multiple rings led to overestimation Overall 62 56 94 94 52 of age. Presumably, identification of accessory a One structure was omitted because readers could not reach agree- growth rings would be more difficult as fish age. ment. Double rings have also been mistakenly identified as separate annuli in dorsal spines of walleye (Ko- Age bias graphs revealed that the first reader covsky and Carline 2000). Conversely, several au- consistently overestimated the age of channel cat- thors (Erickson 1983; Welch et al. 1993; Nash and fish when using cut spines and articulating process Irwin 1999) have reported that spine ages can un- sections; however, his age estimates based on oto- derestimate the age of old fish. Most often, ex- liths were unbiased (Figure 4). Often, annular pansion of the central lumen and the inability to marks on cut spines and articulating process sec- separate annuli near the edge of the structure are tions consisted of multiple growth rings, resulting cited as explanations of missing annuli. Kocovsky in overestimates of age by the first reader. Acces- and Carline (2000) reported difficulty in estimat- sory marks were not observed on otoliths. No bias ing ages from sections of dorsal spines from wall- was noted for the second reader or for age esti- eye because of merged rings, partial rings, and mates based on a concert reading. crowding of distal rings; they attributed such struc- Estimates of mean length at age and age struc- tural abnormalities to slow growth. Accessory ture based on age assignments by the individual marks were not observed in the channel catfish readers often differed from actual mean length at otoliths we examined, and many of the difficulties age and age structure (Table 2). For example, using associated with estimating ages of spines are not cut spines, the first reader estimated that the pop- relevant to otoliths. Therefore, otoliths may pro- ulation consisted of seven year classes when only vide accurate age estimates for channel catfish old- four existed, and the abundance of age-3 fish was er than age 4. drastically underestimated. Following a concert Inconsistencies resulting from variable interpre- reading, otoliths provided estimates of mean tation can lead to improper age assignments, even length-at-age and age structure similar to actual for validated methods of estimating age (Lai et al. values for age-1–4 fish; spines yielded similar val- 1996). Both readers in our study had numerous ues for age-1–3 fish. years of experience determining fish age; however, their relative accuracy differed. Thus, reader ex- Discussion perience may not be a good indicator of reader After concert readings, age assignments were accuracy. Erroneous age estimates, to the extent accurate for age-1–4 channel catfish as determined observed for the first reader when using pectoral by using otoliths and for age-1–3 channel catfish spines, may be sufficient to result in inappropriate
940 BUCKMEIER ET AL. FIGURE 4.—Age bias graphs for reader 1, reader 2, and a joint (concert) reading for age estimates of channel catfish from otoliths, cut pectoral spines, and articulating process (AP) sections. Numbers indicate sample sizes. The dashed line represents agreement between age estimates and known ages. management decisions. The use of multiple readers we believe that sacrificing a sample of fish to pro- and a concert reading in our study reduced reader vide age estimates will not significantly affect bias and provided accurate data. Ideally, individ- most channel catfish populations. In instances uals should verify that they are correctly identi- when fish cannot usefully be sacrificed, cut pec- fying annuli by using known-age fish whenever toral spines may provide an alternative method for possible. However, if known-age fish are not avail- estimating the age of channel catfish of age 3 or able, our results suggest the use of multiple readers younger. However, a sample of otoliths should ini- and a concert reading may ensure data accuracy. tially be used to confirm ages estimated from cut We believe that otoliths are the best structure spines. Regardless of the method, some attempt for estimating age of channel catfish. We did not should be made to verify that annuli are being measure processing time for each method evalu- identified correctly. ated; however, the additional time required to col- Acknowledgments lect and prepare otoliths versus cut spines was rel- Editorial comments by R. Heidinger, S. Van atively small. Both methods were faster than pre- Horn, P. Michaletz, R. Holland, and biologists at paring articulating process sections. In addition, Heart of the Hills Research Station greatly im-
MANAGEMENT BRIEFS 941 TABLE 2.—Comparison of actual and estimated mean Crumpton, J. E., M. M. Hale, and D. J. Renfro. 1987. total lengths (TL) from otoliths, cut pectoral spines, and Aging of three species of Florida catfish utilizing articulating process (AP) sections. Each column represents three pectoral spine sites and otoliths. Proceedings the same group of fish, with the number of age structures of the Annual Conference Southeastern Association in parentheses. of Fish and Wildlife Agencies 38(1984):335–341. Erickson, C. M. 1983. Age determination of Manitoban Estimate walleyes using otoliths, dorsal spines, and scales. Actual Concert North American Journal of Fisheries Management Age mean TL Reader 1 Reader 2 reading 3:176–181. Otoliths Hales, L. S., Jr., and M. C. Belk. 1992. Validation of otolith annuli of bluegills in a southeastern thermal 0 325 (1) 331 (1) 1 321 (13) 318 (15) 320 (12) 321 (13) reservoir. Transactions of the American Fisheries 2 323 (9) 328 (6) 326 (8) 323 (9) Society 121:823–830. 3 368 (14) 374 (15) 364 (15) 368 (14) Hall, D. L. 1991. Age validation and aging methods for 4 452 (3) 424 (1) 452 (3) 443 (2) stunted brook trout. Transactions of the American 5 471 (1) 471 (1) Fisheries Society 120:644–649. Cut spines Heidinger, R. C., and K. Clodfelter. 1987. Validity of 0 311 (1) the otolith for determining age and growth of wall- 1 321 (28) 320 (19) 320 (29) 320 (29) eye, striped bass, and smallmouth bass in power 2 335 (18) 331 (23) 337 (17) 335 (16) plant cooling ponds. Pages 241–251 in R. C. Sum- 3 370 (16) 363 (8) 375 (17) 371 (16) merfelt and G. E. Hall, editors. The age and growth 4 452 (3) 362 (6) 352 (1) of fish. Iowa State University Press, Ames. 5 412 (7) 448 (2) 448 (2) Hining, K. J., J. L. West, M. A. Kulp, and A. D. Neu- 6 325 (1) 462 (1) bauer. 2000. Validation of scales and otoliths for AP sections estimating age of rainbow trout from southern Ap- 1 321 (28) 322 (19) 320 (28) 322 (28) palachian streams. North American Journal of Fish- 2 335 (18) 339 (16) 334 (17) 335 (17) eries Management 20:978–985. 3 370 (16) 332 (11) 370 (15) 369 (13) Kocovsky, P. M., and R. F. Carline. 2000. A comparison 4 380 (6) 374 (1) 372 (3) of methods for estimating ages of unexploited wall- 5 371 (5) eyes. North American Journal of Fisheries Man- 6 326 (3) 7 agement 20:1044–1048. 8 367 (1) Lai, H. L., V. F. Gallucci, D. R. Gunderson, and R. F. Donnelly. 1996. Age determination in fisheries: methods and applications to stock assessment. Pag- proved this manuscript. We thank the staff of the es 82–178 in V. F. Gallucci, S. B. Saila, D. J. Gus- tafson, and B. J. Rothschild, editors. Stock assess- Alabama Cooperative Fish and Wildlife Research ment quantification methods and applications for Unit for collecting fish and removing aging struc- small-scale fisheries. CRC Press, New York. tures. The Alabama Agricultural Experimental Maceina, M. J. 1988. Simple grinding procedure to sec- Station, Auburn University, furnished fish and the tion otoliths. North American Journal of Fisheries pond. The otolith illustration in Figure 2 was Management 8:141–143. drawn by R. G. Howells. Funding was provided Marzolf, R. G. 1955. Use of pectoral spines and ver- by the Alabama Department of Conservation and tebrae for determining age and rate of growth of the channel catfish. Journal of Wildlife Management 19: Natural Resources, Game and Fish Division, 243–249. through Federal Aid in Sport Fish Restoration Mayhew, J. K. 1969. Age and growth of flathead catfish Grant F-40-30 and the Texas Parks and Wildlife in the Des Moines River, Iowa. Transactions of the Department. American Fisheries Society 98:118–120. Muncy, R. J. 1959. Age and growth of channel catfish References from the Des Moines River, Boone County, Iowa, Bagenal, T. B., and F. W. Tesch. 1978. Age and growth. 1955 and 1956. Iowa State Journal of Science 34: Pages 101–136 in T. B. Bagenal, editor. Methods 127–137. for assessment of fish production in fresh waters. Nash, M. K., and E. R. Irwin. 1999. Use of otoliths Blackwell Scientific Publications, London. versus pectoral spines for aging adult flathead cat- Beamish, R. J., and G. A. McFarlane. 1983. The for- fish. Pages 309–316 in E. R. Irwin, W. A. Hubert, gotten requirement for age validation in fisheries C. F. Rabeni, H. L. Schramm, Jr., and T. Coon, ed- biology. Transactions of the American Fisheries So- itors. Catfish 2000: proceedings of the international ciety 112:735–743. Ictalurid symposium. American Fisheries Society, Campana, S. E., M. C. Annand, and J. I. McMillan. 1995. Symposium 24, Bethesda, Maryland. Graphical and statistical methods for determining Prentice, J. A., and B. G. Whiteside. 1975. Validation the consistency of age determinations. Transactions of aging techniques for largemouth bass and channel of the American Fisheries Society 124:131–138. catfish in central Texas farm ponds. Proceedings of
942 BUCKMEIER ET AL. the Annual Conference Southeastern Association of Turner, P. R. 1982. Procedures for age determination Game and Fish Commissioners 28(1974):414–428. and growth rate calculations of flathead catfish. Pro- Sneed, K. E. 1951. A method for calculating the growth ceedings of the Annual Conference Southeastern of channel catfish, Ictalurus lacustris punctatus. Association of Fish and Wildlife Agencies Transactions of the American Fisheries Society 80: 34(1980):253–262. 174–183. Welch, T. J., M. J. Van Den Avyle, R. K. Betsill, and Taubert, B. D., and J. A. Tranquilli. 1982. Verification E. M. Driebe. 1993. Precision and relative accuracy of the formation of annuli in otoliths of largemouth of striped bass age estimates from otoliths, scales, bass. Transactions of the American Fisheries So- and anal fin rays and spines. North American Jour- ciety 111:531–534. nal of Fisheries Management 13:616–620.
You can also read